# [rrd-users] suspicious interpolation

Philippe Charnier charnier at xp11.frmug.org
Sat Oct 11 16:50:30 MEST 2003

Hello,

I'm using rrdtool 1.0.45. I would like to know if the following is the
expected behaviour.

rrdtool create a.rrd --start 999999990 --step 10 \
DS:f:GAUGE:30:U:U RRA:MAX:0.5:1:10
rrdtool update a.rrd 1000000000:2

I put the point at a pdp_st (at 1000000000) so I get the exact value

999999990: nan
1000000000: 2.0000000000e+00
1000000010: nan

rrdtool update a.rrd 1000000009:2

I have not yet crossed next pdp_st (at 1000000010), so my value is not
yet fetchable, but I have 2*9 = 18 to incorporate.

rrdtool update a.rrd 1000000020:4

I crossed both 1000000010 and 1000000020, so I have two values to incorporate:

- from 1000000000 to 1000000010 I have 2*9 + 4*1 = 22
- from 1000000010 to 1000000020 I have 4*10      = 40

So, after dividing by the size of the interval, I would be pleased to find

1000000000: 2.0000000000e+00
1000000010: 2.2000000000e+00
1000000020: 4.0000000000e+00

Unfortunately rrdtool take all of them into one

- from 1000000000 to 1000000020 it counted the same 2*9 + 4*11 = 62
and filled all crossed pdp_st (at 1000000010 and 1000000020) with that value
(after dividing by the size of the whole interval)

1000000000: 2.0000000000e+00
1000000010: 3.1000000000e+00
1000000020: 3.1000000000e+00

This makes 1000000020:3.1 really different from the reality 1000000020:4 and
1000000010:3.1 not so close to 1000000009:2.

Any comment?

------                                                            ------
Philippe Charnier          charnier@{xp11.frmug.org,free.fr,FreeBSD.org}

``a PC not running FreeBSD is like a venusian with no tentacles''
------------------------------------------------------------------------

--
Unsubscribe mailto:rrd-users-request at list.ee.ethz.ch?subject=unsubscribe
Help        mailto:rrd-users-request at list.ee.ethz.ch?subject=help
Archive     http://www.ee.ethz.ch/~slist/rrd-users